Automated mapping of tropical deforestation and forest degradation: CLASlite
Automated mapping of tropical deforestation and forest degradation: CLASlite
复制标题
DOI：
10.1117/1.3223675
复制
发表时间：
2009-08-18
影响因子：
1.7
通讯作者：
Paez-Acosta, Guayana
中科院分区：
文献类型：
--
作者：
Asner, Gregory P.;Knapp, David E.;Paez-Acosta, Guayana
监测森林砍伐和森林退化是评估热带地区碳储存、生物多样性和许多其他生态过程变化的核心。卫星遥感是监测大片地区森林覆盖变化和退化的最准确和最具成本效益的方法，但这些工具和方法都是高度手工操作和耗时的，往往需要专家知识。我们提出了一个新的用户友好的，全自动化的系统CLASlite，它提供了桌面地图的森林覆盖，森林砍伐和森林干扰使用先进的大气校正和光谱信号处理方法与Landsat，SPOT，和许多其他卫星传感器。CLASlite在一台标准的Windows计算机上运行，每小时的处理时间可以绘制超过10，000公里（2）的森林面积，空间分辨率为30米。CLASlite的输出包括活的和死的植被覆盖、裸露土壤和其他基质的百分比图，沿着每个图像像素的不确定性的定量测量。然后使用自动决策树从森林覆盖、毁林和森林干扰的角度对这些地图进行解释。CLASlite输出图像可以直接输入到其他遥感程序、地理信息系统（GIS）、Google Earth（TM）或其他可视化系统。在这里，我们提供了一个详细的描述CLASlite的方法与巴西，秘鲁，和世界各地的其他热带森林网站的森林砍伐和森林退化的情况下的示例结果。
Monitoring deforestation and forest degradation is central to assessing changes in carbon storage, biodiversity, and many other ecological processes in tropical regions. Satellite remote sensing is the most accurate and cost-effective way to monitor changes in forest cover and degradation over large geographic areas, but the tools and methods have been highly manual and time consuming, often requiring expert knowledge. We present a new user-friendly, fully automated system called CLASlite, which provides desktop mapping of forest cover, deforestation and forest disturbance using advanced atmospheric correction and spectral signal processing approaches with Landsat, SPOT, and many other satellite sensors. CLASlite runs on a standard Windows-based computer, and can map more than 10,000 km(2), at 30 m spatial resolution, of forest area per hour of processing time. Outputs from CLASlite include maps of the percentage of live and dead vegetation cover, bare soils and other substrates, along with quantitative measures of uncertainty in each image pixel. These maps are then interpreted in terms of forest cover, deforestation and forest disturbance using automated decision trees. CLASlite output images can be directly input to other remote sensing programs, geographic information systems (GIS), Google Earth (TM), or other visualization systems. Here we provide a detailed description of the CLASlite approach with example results for deforestation and forest degradation scenarios in Brazil, Peru, and other tropical forest sites worldwide.